A modular approach for multi-scale whole-brain simulation: Integrating a neuron-astrocyte-vascular model into The Virtual Brain

Abstract

Whole-brain simulation frameworks have become important tools for studying large-scale brain dynamics, disease mechanisms, and personalized neuroimaging signals such as BOLD fMRI. As these platforms move toward increased biological realism and multi-scale integration, their underlying architecture becomes a critical factor for ensuring scalability, extensibility, and long-term usability. Integrating heterogeneous biophysical processes across neuronal, glial, and vascular domains poses significant architectural challenges. Here, we demonstrate that a modular design is not only compatible with, but fundamentally necessary for, incorporating complex multi-compartment physiological cascades into whole-brain simulation frameworks without modifying their core computational infrastructure. Using The Virtual Brain (TVB) as a reference platform, we present a case study in which a Neuron-Astrocyte-Vascular (NAV) extension is integrated as a set of modular components. In this implementation, astrocytic calcium dynamics and neurovascular coupling mechanisms are incorporated alongside existing neuronal models while preserving core system properties such as computational efficiency, scalability, and personalization. Our results show that neuronal, astrocytic, and vascular subsystems can be implemented as independent yet interoperable layers within the simulation framework, maintaining the integrity of connectivity, coupling, and solver structures. This integration blueprint provides a generalizable template for extending whole-brain models toward richer multiscale representations. To our knowledge, this work represents the first demonstration of integrating a biologically detailed Neuron-Astrocyte-Vascular cascade into a modular whole-brain simulation framework without altering its core architecture. Beyond its immediate application, this approach facilitates more flexible and scalable neuroscientific modeling and opens new opportunities for brain-inspired neuromorphic systems and energy-aware computing.